Latest Tech News
Daily updates from AI, software, hardware and more.
NVIDIA says NIM optimizations lift Nemotron 3 Ultra serving throughput 2.5x
NVIDIA says NIM 2.0.12 raises Nemotron 3 Ultra serving throughput 2.5x on four B200 GPUs for agentic workloads.
Meta details closed-loop liquid cooling for denser AI data centers
Meta says its closed-loop liquid cooling recirculates coolant for AI data centers and cut fan energy 20% in a pilot.
NVIDIA adds NVHBM to NVLink Fusion for custom AI data center chips
NVIDIA added NVHBM to NVLink Fusion, promising more bandwidth, lower HBM power and AWS Trainium collaboration for custom...
Intel details Diamond Rapids, Crescent Island and Wildcat Lake for agentic AI
Intel outlined Diamond Rapids, Crescent Island and Wildcat Lake architectures for agentic AI at Hot Chips 2026.
Meta and NABTU partner on skilled-trades training for AI infrastructure buildout
Meta and NABTU announced a partnership to expand skilled-trades training for workers building AI infrastructure.
Intel proposes $15 billion stock offering to fund AI-era semiconductor growth plans
Intel proposed a $15 billion stock offering to support capital spending and AI-era semiconductor growth opportunities.
Firebird launches Armenia AI factory with plans for 70,000 NVIDIA Rubin and Blackwell GPUs
Firebird launched an Armenia AI factory and plans more than 70,000 NVIDIA Rubin and Blackwell GPUs by 2027.
Microsoft opens new Hyderabad cloud region to expand AI capacity in India
Microsoft opened its India South Central cloud region in Hyderabad, expanding local AI and cloud capacity for enterprise...
NSF launches $100M regional AI infrastructure hubs program for researchers
NSF announced a $100 million regional AI infrastructure hubs program to expand research compute access across the U.S.
Samsung shows zHBM and zNAND-O concepts for denser AI memory at FMS 2026
Samsung presented zHBM and zNAND-O concept memory architectures at FMS 2026 for future AI data center systems.
Google urges session-aware load balancing for real-time AI agents
Google says real-time AI agents need load balancing based on active sessions as well as CPU, memory and request metrics.
OpenAI details GPT-Live architecture for continuous realtime voice AI
OpenAI detailed GPT-Live, a continuous realtime voice architecture that separates audio streaming from reasoning and too...